CHAPTE! THI!TY"TH!EE !e Birth of Grammar in Greece Andreas U. Schmidhauser Grammar as one understands it today gives an account of the system of rules govern- ing the construction of syllables, words, and sentences in a certain language. !e sci- ence thus called was independently—and very di"erently—developed at about the same time in ancient India and Greece: Sanskrit grammar is the work of Pā!ini ( !. c. $%% BC); Greek grammar is the creation of Chrysippus of Soli ( !. &$% BC). Both Pā!ini and Chrysippus not only inaugurate a new 'eld but also represent the culmi- nation of centuries of linguistic thought: what distinguishes them from their prede- cessors is that they, for the 'rst time, integrated the results obtained into one theory. !e term “grammar” itself is of Greek origin: literally, ἡ γραμματική (or in full: ἡ γραμματικὴ τέχνη) is the skill, expertise, or knowledge belonging to a person consid- ered γραμματικός; and the adjective γραμματικός is derived from the noun γράμμα [letter], which in turn derives from the verb γράφειν [write, draw]. Over time, the meaning of γραμματικός and thus also of γραμματική changed. One can distinguish four stages: i. In the fourth century BC, when γραμματικός 'rst appears, it is used to describe someone who knows the “letters”: a person versed in grammar, that is, knows how to read and write, can set apart vowels, consonants, and semiconsonants, and suchlike (e.g. Pl. Cra. $;<e; Phlb. <=d; Soph. &>;a). ii. From the third century BC, γραμματική comes to be used for what one would now call philology and criticism (e.g. Dion. !rax ap. Sext. Emp. Math. <.>?; cf. Di Benedetto &%%?: &.>&&). !us the oeuvre of Aristarchus of Samothrace ( !. <@% BC)—ὁ γραμματικώτατος [the most grammatical] to some (Ath. <>.<&.&)—consists in edi- tions of and commentaries on Homer, Hesiod, Archilochus, Alcman, Pindar, Herodo- tus, and others, as well as in a number of critical treatises on Homeric questions (see e.g. Pfei"er <D@=: &<%). iii. From the early 'rst century BC, the content of the grammarians’ discipline is en- larged; in particular, it includes a new so-called technical [τεχνικόν] part or tool, which corresponds to what one would now term grammar (e.g. Asclepiades ap. Sext. Emp. Math. <.D<; <.&>&; cf. Blank <DD=: <$@). A “technical grammarian” such as Apol- 2 ANDREAS U. SCHMIDHAUSER lonius Dyscolus ( !. AD <$%) no longer edits texts nor comments on them—he com- poses treatises on the elements, the parts of speech, etc. (Suda α ;$&&; cf. Ap. Dy. Con. &<;.<%). iv. From the end of antiquity, the τεχνικόν progressively comes to be seen as the γραμματική par excellence (e.g. Michael Sync. Synt. $@ [c. =<;]). !us Priscian ( !. >%%), the author of, inter alia, a voluminous and immensely inFuential Latin gram- mar, can now refer to Apollonius as summus artis auctor grammaticae [the greatest authority on grammatical science] (Inst. =.$;D.&&). It is this use of γραμματική which has given rise to the modern notion of grammar. (Two precisions to the above quadripartition: 'rst, the use of the term evolved gradually, of course; and secondly, the creation of a new use did not, in this case, en- tail that the older uses passed away entirely.) Now some might argue the semantic shiG just outlined makes it impossible to write a historiography of Greek grammar from Plato to, say, Planudes ( !. <;%%) inso- far as there does not exist a single discipline called γραμματική, the history of which one could study. Yet to renounce the project entirely would be rash. For there still remains the possibility of focusing on one relatively stable acception of the term γραμματική, and studying the history of that discipline. Furthermore, one ought not to forget that past nomenclatures are immaterial to the question whether the inquiry is the same as, or similar to, the one practised later. Hence, if one intends to study the history of grammar qua science of language, one should not want to restrict oneself to studying the history of the τεχνικόν and of γραμματική in its last use. For it is well known, and I shall show below, that the subject as such was 'rst recognised by the Stoics. Because the Stoics’ inquiry was done under the umbrella of philosophy, it is oGen declared—generally without further argument—that theirs was not yet an “autono- mous science” (e.g. Di Benedetto &%%?: &.$D?). Some scholars have even gone so far as to argue that philosophy “blocked” the emancipation of grammar (e.g. Ildefonse <DD?: <>). Both claims are anachronistic and false, for they presuppose that philoso- phy and science are radically di"erent in nature—which, at least in antiquity, they were not: any historiography of biology, for example, starts with Aristotle, who in- vented the discipline (see e.g. Lennox &%%<: xx). As for the alleged lack of autonomy, depending on how one understands the notion, this need not necessarily be a point of criticism: aGer all, for half a century now most linguists who reFect on such ques- tions have considered their discipline a branch of cognitive science (e.g. Chomsky <D@=: <). Grammar—and from now on, I shall use that term to speak of linguistic sci- ence—was one of the pillars of education in Antiquity and the Middle Ages. !us one encounters examples of grammatical analysis in ancient texts of all genres—from rhetoric and philosophy to medicine and theology. Its inFuence went far beyond the THE BIRTH OF GRAMMAR IN GREECE 3 Greek world. Latin, in the late second century BC, became the 'rst language to which the Greek system was adapted; and for the next six hundred years, Latin grammari- ans continued to be inspired by their Greek colleagues (the reverse does not hold). In the sixth and seventh centuries, the Techne (a brief school grammar) and other works were translated into languages as diverse as Syriac, Armenian, and Georgian. In each case, the resulting grammars were the 'rst for that language. And so it spread. !e Greek heritage, then, was vast. One is so used to speaking of nouns and pronouns, of the 'rst person and the past tense, of case and gender, that it may sometimes prove diHcult to remember that these are all technical notions we inherited from antiquity. !ey may—or may not— have been appropriate for describing Ancient Greek. Yet to decide whether they are suited to other languages, or to language in general, one 'rst has to know what they meant, and why they were introduced. !e study of the origins of the study of lan- guage is thus indispensable to the study of language itself. !e Beginnings Some situate the genesis of grammar in the Dark Ages, regarding the creation of the Greek alphabet, in the late ninth century BC, as “the 'rst achievement of linguistic scholarship in Greece” (Robins <DD?: <@). !is is confused on two counts: First, on such an inclusive conception of linguistic scholarship, one surely ought to start at least seven hundred years earlier, for the creation of a syllabary—Linear B (see Chap- ter <)—requires linguistic pro'ciency, too—indeed, every script presupposes some linguistic analysis. Secondly, the changes in respect to the Phoenician consonantal alphabet are but few: the graphemes for the glottal and pharyngeal obstruents /! "/ —which do not occur in Greek—were set to represent the vowels /a o/; and with the grapheme for /ħ/ already assigned to Greek /h/, the grapheme for /h/ was set to rep- resent the vowel /e/. Otherwise the Greek alphabet closely resembles its model; even the order of the letters and their names are the same (see Burkert &%%>: &D$). In the 'Gh and fourth centuries BC, a variety of linguistic phenomena were for the 'rst time identi'ed and labeled and, sometimes, explained. !e Sophists, in par- ticular, appear to have shown an intense interest in language. (!eir writings on the subject have been lost, so that one is dependent on later authors for information.) Protagoras ( !. $>% BC), the most celebrated of that heterogeneous group, “divided up the kinds of names [τὰ γένη τῶν ὀνομάτων]: male [ἄρρενα], female [θήλεα], and inanimate [σκεύη]” (Arist. Rh. <$%?b@); “distinguished the parts of time [μέρη χρόνου]” (Diog. Laert. D.>&); and “divided speech [λόγος] into four kinds: prayer, question, answer, order” (ibid. D.>;). !ere can be little doubt that one has here the ancestors of the grammatical categories of gender, tense, and sentence. 4 ANDREAS U. SCHMIDHAUSER Plato ( !. ;=% BC) is the most important 'gure in the prehistory of grammar. On every linguistic level—element, syllable, word, sentence—the distinctions he draws, the terms he introduces, the arguments he advances (and also those he thinks he refutes) have leG their imprint on the Stoic and thus the Apollonian system. An illus- tration—one inFuential passage—must here suHce: When, in the Sophist, he analyses λόγος, he distinguishes between names [ὀνόματα] and verbs [ῥήματα], characteris- ing them in semantic terms: a verb is “an indication applied to actions [ἐπὶ ταῖς πράξεσιν]”, whereas a name is “a vocal sign applied to those performing them” (&@&a). To say something, then, one cannot just utter a list of names or of verbs: the smallest sentence [λόγος] is a combination [συμπλοκή] of a name and a verb (&@&b– d). !is brilliant—and now seemingly trivial—insight permits Plato to provide a pre- cise account of truth and falsehood: to be true, a (simple) sentence must say of “what is that it is [τὰ ὄντα ὡς ἔστιν]”: Θεαίτητος νεῖ, for instance, is false insofar as what is said or predicated of !eaetetus is not something that he is—for !eaetetus is not swimming now, but talking to the Stranger (&@;a–d; cf. Frede <DD&: $<&).